![](https://media.wired.com/photos/5f7d13ce941317efa4fcd4f8/191:100/w_1280,c_limit/Science-covid-diabetes-94497712.jpg)
Take for example, three men who showed up at a hospital in the northern part of India weak, feverish, and without any history of diabetes. They all tested positive for SARS-CoV-2. And when their bloodwork came back, they all had dangerously high buildups of glucose and ketones, which the body produces when it doesn’t have enough insulin to break down sugar. The official term for the potentially deadly complication is diabeteic ketoacidosis, and it is usually seen in children with type 1 diabetes.
Mohammad Shafi Kuchay, an endocrinologist who consulted on the cases, told WIRED via email that he and the other doctors assigned to the cases assumed the virus had somehow knocked out these patients’ insulin-making cells, giving them type 1 diabetes. And so the doctors put the men on a regimen of insulin injections. But as the months went by, they needed the injections less and less. They were shifted to oral antidiabetic drugs, and have been managing like this for more than two months now. “That means the patients have type 2 diabetes,” Shafi Kuchay wrote. Whatever damage the virus caused to these patients’ insulin-secreting cells appears to be transient. More monitoring will be necessary to determine if their type 2 diabetes diagnosis is short-lived, as well.
Will other patients also find that their blood sugar problems go away when their infection resolves? Or will Covid-19 cause diabetes for life? “None of those questions could be addressed with just the anecdotal case reports that were coming out,” says Rubin. That’s why he and an international group of scientists decided to act on their hunch and establish a global repository for tracking information about the coronavirus-related diabetes.
The CoviDiab Registry allows doctors around the world to upload anonymized data about Covid-19 patients with abnormal blood sugar levels who have no prior history of diabetes. That includes the basics, like age, sex, past medical history, and details of the person’s disease progression: Do they wind up in the ICU or on a ventilator? But the registry also asks doctors to catalogue which medications they’re administering, on the off chance that it’s not the coronavirus infection that’s triggering the diabetes but an unknown side effect of something used to treat it.
The goal of this information-gathering effort is to grok the scale and scope of the problem, as well as potential solutions. How often is Covid-19 associated with new-onset diabetes? And what flavor—type 1, type 2, or a new form of the disease? What exactly causes the metabolic malfunction? How long do such cases of diabetes last, and what are the best ways to treat them? It might be a while before there’s enough data to answer questions about prevalence and mechanism. But Rubin thinks they could have information about what kinds of diabetes most frequently develop in Covid-19 patients—and what might predispose people to this particular coronavirus complication—by the end of the year.
This is not the first registry to track the overlap between people with Covid-19 and other conditions. Similar data pooling efforts spun up early this year for people with inflammatory bowel disease, chronic liver disease, and rheumatoid disorders, among others. Eric Topol, a leading cardiologist at the Scripps Research Translational Institute, has advocated for a similar approach to tracking heart complications of Covid-19. The diabetes database launched in June, and since then, more than 275 physicians have requested access to share data about at least one patient who meets the criteria. Vetting each physician contributor takes time, and after that, registry organizers have to set up data-sharing agreements that comply with Europe’s strict data protection laws. But so far, dozens of cases have been uploaded. And what’s clear, says Rubin, is that this is not just an anecdotal problem. “From what we’ve seen so far, Covid-19-related diabetes will not be a prevalent issue that affects the majority of people,” he says. “But now we know it’s a possibility, even if not a common one.”